KR102364951B1 - Systems and methods for improving inspection sensitivity of inspection tools - Google Patents

Abstract

Systems and methods for improving inspection sensitivity for detecting defects in wafers using inspection tools are disclosed. A plurality of light emitting diodes illuminate at least a portion of the wafer to capture a set of grayscale images. A residual signal is determined from each image in the grayscale image set, and the residual signal is subtracted from each image in the grayscale image set. Defects are identified based on the subtracted grayscale image set. Models of inspection tools and wafers can be fabricated and refined with some embodiments of the disclosed systems and methods.

Description

Systems and methods for improving inspection sensitivity of inspection tools

This application claims priority to and pending US Provisional Application No. 62/133,959, filed March 16, 2015, the contents of which are incorporated herein by reference.

The present disclosure relates to systems and methods for improving the inspection sensitivity of an inspection tool, such as a metrology tool.

Wafer inspection systems help semiconductor manufacturers increase and maintain integrated circuit (IC) chip yields. The IC industry uses inspection systems to detect defects that occur during the manufacturing process. Its main purpose is to monitor whether a process is under control. If the process is outside the scope of the established norm, the system shall indicate the problem and/or the cause of the problem, which can be corrected by the manager of the IC manufacturing process. Some important inspection system characteristics are defect detection sensitivity and wafer throughput. When sensitivity and throughput are combined, higher sensitivity usually lowers throughput. There are physical and economic reasons for this relationship.

The relative values of sensitivity and throughput depend on the capabilities of the inspection system. There are three general functional requirements for this system: first, in process development; Second, in the monitoring of the process line; Third, in the monitoring of the station, it is the detection and classification of defects. In process development, low throughput can be tolerated to capture small defects and a wider range of defect types. However, in the monitoring of a production line or station, it becomes relatively important in terms of cost-of-ownership and throughput. In this case, of course, the sensitivity should be suitable for capturing yield limiting defects.

The evolution of the semiconductor manufacturing industry is placing increasing demands on yield management, particularly metrology and inspection systems. Critical dimensions are shrinking while wafer sizes are increasing. The economy is leading the industry in reducing the time it takes to achieve high-margin, high-value-added production. Therefore, minimizing the total time it takes to detect and resolve yield problems determines the return on investment for semiconductor manufacturers.

Accordingly, inspection systems are evolving from stand-alone "tools" for finding defects to become part of a more complete solution that performs the functions of defect detection, defect classification, result analysis, and corrective action recommendations.

Existing systems and methods have been used for automatic defect inspection of semiconductor wafers. However, the inspection parameters of prior art systems and methods are rather limited in high-throughput environments. For example, parameters such as coated film thickness or process uniformity across the wafer are time consuming and computationally expensive.

Current systems capture grayscale images of semiconductor wafers under all possible combinations of red, green and blue LED illumination. Grayscale images are currently used for automatic defect detection of semiconductor wafers or to detect wafer-to-wafer process variation (G-view).

One embodiment of the present disclosure is a method for identifying defects in a wafer with an inspection tool. The method includes capturing a set of grayscale images of the wafer by using an electronic image capture device of an inspection tool. The set of grayscale images includes illuminating at least a portion of the wafer with blue wavelength light to capture a first grayscale image, illuminating at least a portion of the wafer with red wavelength light to capture a second grayscale image, and green illuminating at least a portion of the wafer with wavelength light to capture a third grayscale image. In an embodiment, the method may further comprise illuminating at least a portion of the wafer with a combination of blue, red, or green wavelength light to capture one or more additional grayscale images.

The method may further comprise storing the set of grayscale images in a computer readable memory.

The method further comprises determining, using a processor in communication with a computer readable memory, a residual signal in respective images of the grayscale image set based on a combination of images in the grayscale image set. may include The residual signal is generated by: building, using the processor, a rigorous mathematical model of defect detection using an inspection tool; determining, using the processor, one or more model parameters based on a known standard grayscale image set (such as a VLSI thin film standard image set); constructing, using the processor, a model of the wafer using one or more model parameters, the model being based on design values or previously measured values (model parameters, etc.); predicting grayscale signals by using a model of the wafer and a rigorous mathematical model; adjusting one or more parameters of the model of the wafer until a best match is found between the measured and predicted grayscale signals from the wafer; reporting, using the processor, one or more parameters corresponding to best match models as measured sample parameters; calculating, using the processor, a residual signal based on differences between the measured grayscale and the predicted grayscale on the wafer; and storing the calculated residual signal in a computer readable memory for future defect detection.

The method may further include, using the processor, subtracting, using the processor, a residual signal of each image of the set of grayscale images from each image of the set of grayscale images.

The method may further include, using the processor, identifying defects in the wafer based on the subtracted set of grayscale images.

The method may further include converting the set of grayscale images captured by the image capture device using an analog-to-digital converter.

The method may further comprise importing wafer information into a computer readable memory, wherein calculating a residual signal in each of the images of the grayscale image set is also based on the imported wafer information. The wafer information may be in GDSII format. The wafer information may be automatically imported by the processor.

The method includes using an electronic image capture device of an inspection tool to capture an additional set of grayscale images of the wafer after the wafer has been modified; determining, using a processor in communication with the computer readable memory, a residual signal of each of the images in the additional grayscale image set based on the combination of images in the additional grayscale image set; subtracting, using the processor, a residual signal of each image of the additional grayscale image from each image of the additional grayscale image set; and identifying, using the processor, a defect in the wafer based on differences between the grayscale image sets.

One embodiment of the present disclosure may be described as an improved inspection tool system. The system may include a control processor and an electronic image capture device in electronic communication with the control processor. The system may further include a plurality of light emitting diodes each configured to emit light of a different wavelength. The plurality of light emitting diodes may be in electronic communication with the control processor.

The system can further include a computer readable memory in electronic communication with the image capture device and an analysis processor in electronic communication with the computer readable memory. In an embodiment, the system may further comprise an analog-to-digital converter configured to convert the set of grayscale images for storage in a computer readable memory.

The control processor may be configured to instruct the plurality of light emitting diodes to illuminate at least a portion of the wafer with blue, red, and green wavelength light to capture first, second, and third grayscale images.

The control processor may also be configured to instruct the electronic image capture device to capture a set of grayscale images of the wafer. Each image of the captured set may be made while at least a portion of the wafer is illuminated by a plurality of light emitting diodes. The control processor may also be configured to store the set of grayscale images in the computer readable memory. The control processor may also be configured to instruct the plurality of light emitting diodes to illuminate at least a portion of the wafer with a combination of blue, red, and green wavelength light to capture an additional grayscale image under the combined light.

The analysis processor may be configured to determine a residual signal in each image of the set of grayscale images retrieved from the computer readable memory based on the combination of images in the set of grayscale images. The analysis processor includes: building, using the analysis processor, a rigorous mathematical model of defect detection using an inspection tool; determining, using the analysis processor, one or more model parameters based on a known set of standard grayscale images; constructing, using the analysis processor, a model of the wafer using one or more model parameters, the model being based on design values or previously measured values; predicting grayscale signals by using a model of the wafer and a rigorous mathematical model; adjusting one or more parameters of the model of the wafer until a best match is found between the measured and predicted grayscale signals from the wafer; reporting, using the analysis processor, one or more parameters corresponding to best match models as measured sample parameters; calculating, using the analysis processor, a residual signal based on differences between the measured grayscale and the predicted grayscale on the wafer; and storing the calculated residual signal in a computer readable memory for future defect detection to determine the residual signal in each image of the grayscale image set.

The analysis processor may also be configured to subtract a residual signal of each image of the grayscale image from each image in the set of grayscale images, and identify defects in the wafer based on the subtracted set of grayscale images. The analysis processor may also be configured to import wafer information from the computer readable memory and determine a residual signal in each of the images of the grayscale image set further based on the imported wafer information. The wafer information may be in GDSII format.

The control processor may also be configured to direct the electronic image capture device to capture an additional set of grayscale images after the wafer is modified. In such an embodiment, the analysis processor also determines a residual signal in each image of the additional grayscale image set based on the combination of images in the additional grayscale image set, and from each image in the additional grayscale image set. Subtract the residual signal of each image in the additional grayscale image sets and identify defects in the wafer based on differences between the grayscale image sets.

For a complete understanding of the features and objects of the present disclosure, reference is made to the following detailed description in conjunction with the accompanying drawings.
1 is a diagram illustrating detection of a bond on a wafer using a system or method of the present disclosure.
2 is a diagram illustrating detection of process drift from wafer to wafer using a system or method of the present disclosure.
3A-3C are exemplary gray level images of red illuminated (FIG. 3A), green illuminated (FIG. 3B), and blue illuminated (FIG. 3C) of a semiconductor wafer using a system or method of the present disclosure.
4A is a diagram illustrating a wafer map of thickness measured using a system or method of the present disclosure.
4B is a diagram illustrating a comparison of the data captured in FIGS. 4A and 4C .
4C is a diagram illustrating a wafer map of thickness measured using a metrology tool.
5A-5C are diagrams illustrating the residual signal of the red, green, and blue channels after the signal from the thickness change is reduced using the system or method of the present disclosure.
6 is an exemplary imaging device used in a system or method of the present disclosure.
7 is a flowchart illustrating an exemplary embodiment of the present disclosure.

While the claimed subject matter will be illustrated by specific embodiments, other embodiments are within the scope of the present disclosure, including embodiments that do not provide all the advantages and features described herein. Various structural, logical, process steps, and electronic changes may be made without departing from the scope of the present disclosure.

Embodiments of the systems and methods disclosed herein enable quantitative monitoring of sample parameters and provide improved test performance. The system produces a more reliable and measurable quantity for each point on the wafer for each wavelength. This increases the possible applications and improves the results. Extracting sample parameters from inspection tools can help detect process parameter drift, which semiconductor manufacturers can take preventive or corrective action.

As used herein, the term “wafer” generally refers to a substrate formed of a semiconductor or non-semiconductor material. Examples of such semiconducting or non-semiconducting materials include, but are not limited to, single crystal silicon, gallium arsenide, indium phosphide, sapphire, and glass. Such substrates may generally be found and/or processed in semiconductor manufacturing facilities.

A wafer may include one or more layers formed on a substrate. For example, such layers may include, but are not limited to, photoresist, dielectric material, conductive material, and semiconductor material. Many different types of such layers are known, and the term wafer as used herein is intended to include wafers comprising all these types of layers.

One or more layers formed on the wafer may or may not be patterned. For example, a wafer may include a plurality of dies each having repetitively patterned features or periodic structures. The formation and processing of these layers of material can ultimately result in finished devices. A number of different types of devices may be formed on a wafer, and the term wafer as used herein is intended to include the wafer from which any type of device known in the art is fabricated.

Embodiments of the disclosed systems and methods extract information from existing data in a mathematically rigorous manner. In one embodiment, the grayscale signal from the bright field channel 603 of the system of FIG. 6 is a portion or all of a wafer with any combination of a predetermined illumination source, ie, red, green and blue LEDs. obtained through a scan across the wafer area. The system of FIG. 6 is calibrated using previous data analysis. A calibration process may be performed by analyzing a grayscale signal as the system scans one or more wafers having a known surface structure, such as a VLSI thin film standard. This calibration process finds all parameters for a rigorous mathematical model of the system of FIG. 6 . During data analysis on the wafer under test, in addition to rigorously modeling the structures on the wafer, a model with predetermined parameters from the calibration process is used. The results of the rigorously modeled sample and system of FIG. 6 are predicted grayscale signals. The predicted grayscale signals are compared with the grayscale signal measured by the system. An optimal match between the predicted signal and the measured signal is obtained by adjusting parameters of the samples under test, such as the thickness of the film, the optical constant of the material, and/or the critical dimension (CD) of some patterned features. The parameters giving the best match are reported as measurement results.

In one embodiment, a rigorous modeling method is applied to an automatic optical inspection device such as that shown in FIG. 6 . As the method improves the performance of the inspection system, the grayscale signal (as illustrated in FIGS. 3A-3C ) becomes increasingly precise. By rigorously analyzing or modeling grayscale data, embodiments of the present disclosure can use grayscale data from an optical inspection device to measure sample parameters, such as the thickness of a thin film stack. 4A illustrates the thickness of a semiconductor wafer as determined using the system and method of the present disclosure via the calibration and data analysis processes disclosed herein. Figure 4c shows the thickness of the same semiconductor wafer as determined by a well-established accurate but time-consuming metrology tool. 4B illustrates the accuracy of the presently disclosed system and method compared to metrology tools by plotting the results from the two methods diagonally. For clarity, the differences between them are shown as squares in FIG. 4B .

In another embodiment, the defect detection sensitivity can be increased by applying a rigorous modeling method to the original grayscale imaging. In one embodiment, an example brightfield RGB grayscale image, such as the images of FIGS. 3A-3C , may be captured. An RGB grayscale image may be a set of grayscale images of a wafer or portion of a wafer illuminated by red, green, or blue light. Each image in this set may correspond to an image captured under a different color of light. A major component of the grayscale variation across the imaged wafer is due in part to the wafer's film thickness variation. After mathematically removing the signal caused by the film thickness change, a residual signal as shown in Figs. 5A-5C can be found.

For example, there are three main data components shown in Figures 5a-5c: (a) an anomalous feature in the ring between the radius range 110-150 mm normalizes the optical properties of the film, especially in the upper left range. process non-uniformities that cause them to behave differently from values; (b) ring-type and horizontal strip-type features are known hardware limitations in the exemplary embodiment; (c) After removing the hardware signature described in (b), the image reveals the weak signal defect with a sub digitized-count. Therefore, the defect detection sensitivity of the system will be improved after removing the major thickness change components.

In other embodiments, the capabilities of the present disclosure may be extended by collecting and analyzing multiple sets of grayscale images acquired on the same wafer but at different times during the wafer process. For example, one set of grayscale images may be captured after each film layer deposition process. In one embodiment, the set of grayscale images may be taken after a pre-lithography layer, an ARC layer, and a photoresist layer. Another set of grayscale images can be taken after developing the pattern. When all sets of grayscale images are analyzed together, the CD value of the patterned structure as well as the thickness of all films can be determined. In addition, after removing major signal components due to film thickness and CD variations from sets of original RGB grayscale images, process variations and small defects can be detected at all process steps with high sensitivity.

In other embodiments the film stack and/or patterned structure information can be imported for stringent analysis. For example, the stack and/or patterned structure information at the location of interest may be imported from a GDSII file or other suitable type of file. Information can be imported automatically or manually. Film stacks and/or patterned structures may have different responses to incident, angle and azimuth of light, numerical aperture, wavelength, polarization, and the like. Variations in the unique film stacks and/or patterned structures within a die or field, within a wafer, and between wafers may reflect changes in the wafer fabrication process. These changes may be detected, isolated and/or decoupled by applying algorithms such as smart image analysis and/or algorithms that rigorously model the system. For example, after the main component of the grayscale change through the thickness change of FIGS. 3A to 3C is subtracted, the residual grayscale change may be determined as shown in FIGS. 5A to 5C . The outer ring in FIG. 5A exhibits high residual and can be identified because the film optical properties differ from other regions during wafer processing. This can be a process tool defect that causes material property changes and goes undetected without removing major grayscale changes with thickness changes. The same algorithm can be applied to discover more types of defects related to a process or process tools.

6 is a diagram of one type of hardware used to capture grayscale images of a wafer. 6 illustrates an embodiment of a wafer inspection system where maximum flexibility with respect to illumination spectra is desired. The wafer inspection system shown in FIG. 6 includes a wafer 600 , objective lenses, a turret, a bright-field illuminator 607 , an illumination relay optics 609 , and an auto focus unit 611 . ), a beamsplitter, a tube lens, a review camera 601 , and a dark field illumination device 613 . To achieve spatial separation, the fields of view of detectors 603 and 605 must fit within the field of view of the objective lens without overlap. The use of linescan CCD or time delay and integration (TDI) CCD sensors facilitates achieving this goal as these sensors have long and thin footprints. However, brightfield and darkfield detectors are not limited to linescan CCD or TDI CCD sensors, and may alternatively be implemented with any other suitable sensor. While using a TDI detector is one example embodiment, there are a number of advantages to using TDI detectors. By using a TDI detector, a system as shown in Fig. 6 can drive a scan of the wafer for a continuously moving detection field while taking a grayscale signal, and a swath (with a length limited only by the wafer size) swaths) can be printed. This can be important when Power-Spectral-Density (PSD) is required from measurements. PSD is a Fourier transform of a measured quantity. To cover a wide spatial frequency range of the measured PSD from one measurement, the use of a small pixel size extending the high frequency end and a long measurement length extending the low frequency end may be desirable. The ratio of full scan to pixel size gives the total number of pixels along the scan direction. For TDI, this can be greater than 1,000,000 (eg, a 300 mm long line with 0.3 micro pixel size), which provides six times the spatial frequency range during one TDI scan. In contrast, for strobe technology, this ratio is limited by the number of pixels in the detector, which is typically less than 2,000 for automated optical inspection tools in the semiconductor industry. This means that the TDI scheme can provide a wide spatial frequency coverage range of more than 500 times in one measurement. Other benefits of using TDI include faster and higher resolution.

The reflected light and scattered light collected by the objective lens are converged into real images by the tube lens. In one embodiment, the brightfield and darkfield images may be separated into appropriate detection channels by a prismatic double-sided mirror. However, there are many other suitable optical components that can be used to separate brightfield and darkfield images.

In one embodiment, the darkfield image is focused directly on the darkfield detector 605 . On the brightfield side, most of the brightfield light is focused on the brightfield detector 603 . However, a small portion of the brightfield light may be split off by a cube beamsplitter and directed to the review camera 601 . A review camera 601 may be used to acquire color images of the swatch under inspection. In some cases, additional optical elements may be disposed between the beamsplitter and review camera 601 to adjust image magnification according to imaging requirements.

A beamsplitter, optical element, and review camera 601 may not be included in all embodiments of the present invention. Once removed, the brightfield image from the double-sided mirror can be focused directly on the brightfield detector 603 . It is also worth noting that a beamsplitter, optical element, and review camera 601 may be added to other embodiments including wide field inspection.

Output signals from the brightfield and darkfield detectors may be passed to a computer (not shown) for further processing. Because the two channels are spatially separated, the brightfield and darkfield detectors can acquire brightfield and darkfield images of the wafer substantially simultaneously. This improves throughput and increases sensitivity to a wider range of defects by allowing the detector output signals to be combined before a defect is determined (compared to a system that can only provide one mode at a time). In addition to brightfield and darkfield defects, the output signals can be combined to locate defects that can only be detected in a brightfield difference versus darkfield difference decision space.

The output signals from the two detectors may be fed to one or more computer systems (not shown) for further processing. For example, the output signals may be supplied to a processor (not shown). The processor may be coupled to the two detectors by a transmission medium (not shown). The transmission medium may include any suitable transmission medium known in the art. The processor may also be coupled to the detector by one or more electronic components (not shown), such as an analog-to-digital converter. In this manner, the processor may be configured to receive output signals from the detectors.

In some embodiments, the processor may be configured to use the output signals to detect one or more defects on the specimen. Defects may include any defects of interest on the specimen. In addition, the processor may be configured to perform any other inspection related functions known in the art (eg, defect location determination, defect classification, defect mapping, etc.). A processor may take a variety of forms, including a personal computer system, a mainframe computer system, a workstation, an image computer, a parallel processor, or any other processing device known in the art. In general, the term “computer system” may be broadly defined to encompass any device having one or more processors that execute instructions from a memory medium. The processor may be configured to use the output signals generally and any method and/or algorithm known in the art to detect defects on the specimen.

7 is a flowchart of an embodiment of the present invention. This particular embodiment is a method that includes capturing 701 a set of grayscale images of a wafer by using an electronic image capture device of an inspection tool. The set of grayscale images is formed by illuminating at least a portion of the wafer with blue wavelength light to capture a first grayscale image (703), illuminating at least a portion of the wafer with red wavelength light to capture a second grayscale image; 705 , and illuminating at least a portion of the wafer with green wavelength light to capture a third grayscale image 707 .

The method may further include converting (709) the set of grayscale images captured by the image capture device using an analog-to-digital converter. The method may further comprise storing (711) the set of grayscale images in a computer readable memory.

The method includes determining (713) a residual signal in respective images of a grayscale image set based on a combination of images in the grayscale image set, using a processor in communication with a computer readable memory. ) may be further included. The residual signal is generated by: building 715, using the processor, a rigorous mathematical model of defect detection using an inspection tool; using the processor, determining ( 717 ) one or more model parameters based on a known standard grayscale image set (such as a VLSI thin film standard image set); using the processor, building 721 a model of the wafer using one or more model parameters, the model being based on design values or previously measured values (model parameters, etc.); predicting (723) grayscale signals using the analysis processor, using the model of the wafer and the rigorous mathematical model; adjusting 725 one or more parameters of the model of the wafer until a best match is found between the measured grayscale signals from the wafer and the predicted 723 grayscale signals; reporting, using the processor, one or more parameters corresponding to the best match models as measured sample parameters (727); using the processor, calculating (729) a residual signal based on differences between the measured grayscale and the predicted grayscale on the wafer; and storing the calculated residual signal in a computer readable memory for future defect detection ( 731 ).

The method may further comprise importing (719) the wafer information into a computer readable memory, wherein calculating a residual signal in each of the images of the grayscale image set also includes: based on The wafer information may be in GDSII format. The wafer information may be automatically imported 719 by the processor. The method may further include, using the processor, subtracting 733 a residual signal of each image of the set of grayscale images from each image of the set of grayscale images. The method may further include, using the processor, identifying ( 735 ) a defect in the wafer based on the subtracted set of grayscale images.

While the claimed subject matter will be illustrated by specific embodiments, other embodiments are within the scope of the present disclosure, including embodiments that do not provide all the advantages and features described herein. Various structural, logical, process steps, and electronic changes may be made without departing from the scope of the present disclosure.

Embodiments of the systems and methods disclosed herein enable quantitative monitoring of sample parameters and provide improved test performance. The system produces a more reliable and measurable quantity for each point on the wafer for each wavelength. This increases the possible applications and improves the results. Extracting sample parameters from inspection tools can help detect process parameter drift, which semiconductor manufacturers can take preventive or corrective action.

In some embodiments, the inspection systems disclosed herein may be configured as “stand-alone tools” or tools that are not physically connected to a process tool. In other embodiments, the inspection systems disclosed herein may be coupled to a process tool (not shown) by a transmission medium that may include wired and wireless portions. The process tool may include any process tool known in the art, such as a lithography tool, an etching tool, a deposition tool, an abrasive tool, a plating tool, a cleaning tool, or an ion implantation tool. A process tool may be configured as a cluster tool or as multiple process modules connected by a common handler. Alternatively, the inspection systems disclosed herein may be integrated into a process tool as described above. In some cases, results of tests performed by the systems disclosed herein may be used to change parameters of a process or process tool using feedback control techniques, feedforward control techniques, and/or in situ control techniques. there is. Parameters of a process or process tool may be changed manually or automatically.

Embodiments of the present disclosure may allow sample parameters from an inspection tool to be extracted and process parameter drift detected at an early stage to allow for preventive action. By doing so, the value of the inspection tool can be increased without significant cost.

While the present disclosure has been described with respect to one or more specific embodiments, it will be understood that other embodiments of the present disclosure may be made without departing from the spirit and scope of the present disclosure. Accordingly, this disclosure is to be considered limited only by the appended claims and a reasonable interpretation thereof.

Claims (17)

A method for identifying defects on a wafer by an inspection tool, comprising:
using an electronic image capture device of an inspection tool to capture a set of grayscale images of the wafer;
illuminating at least a portion of the wafer with blue wavelength light to capture a first grayscale image;
illuminating at least a portion of the wafer with red wavelength light to capture a second grayscale image; and
capturing a set of grayscale images of the wafer by illuminating at least a portion of the wafer with green wavelength light to capture a third grayscale image;
storing the set of grayscale images in a computer readable memory;
determining, using a processor in communication with the computer readable memory, a residual signal in respective images of the grayscale image set based on a combination of images in the grayscale image set;
subtracting, using the processor, the residual signal of each image of the set of grayscale images from each image of the set of grayscale images; and
identifying, using the processor, defects in the wafer based on the subtracted set of grayscale images;
A method for identifying defects in a wafer by an inspection tool, comprising: According to claim 1,
wherein capturing the set of grayscale images of the wafer further comprises capturing one or more additional grayscale images by illuminating at least a portion of the wafer with a combination of blue, red, or green wavelength light. How to identify defects in wafers by According to claim 1,
and converting the set of grayscale images captured by the image capture device using an analog-to-digital converter. The method of claim 1,
Determining a residual signal in each of the images of the grayscale image set comprises:
building, using a processor, a rigorous mathematical model of defect detection using the inspection tool;
determining, using the processor, one or more model parameters based on a known set of standard grayscale images;
constructing, using the processor, a model of the wafer using the one or more model parameters, the model being based on design values or previously measured values;
predicting grayscale signals by using the model of the wafer and the rigorous mathematical model;
adjusting one or more parameters of the model of the wafer until an optimal match is found between the predicted grayscale signals and the grayscale signals measured from the wafer;
reporting, using the processor, the one or more parameters corresponding to the model for which the best match was found as measured sample parameters;
calculating, using the processor, a residual signal based on differences between the predicted grayscale and the measured grayscale of the wafer; and
storing the calculated residual signal in a computer readable memory for future fault detection;
A method for identifying defects in a wafer by an inspection tool, comprising: 5. The method of claim 4,
wherein the known standard grayscale image is a VLSI thin film standard image set. 5. The method of claim 4,
wherein the previously measured values are model parameters. According to claim 1,
further comprising importing wafer information into the computer readable memory;
and determining a residual signal in each of the images of the set of grayscale images is further based on the imported wafer information. 8. The method of claim 7,
wherein the wafer information is in GDSII format. 8. The method of claim 7,
and the wafer information is automatically imported by the processor. According to claim 1,
using the electronic image capture device of the inspection tool to capture an additional set of grayscale images of the wafer after the wafer has been modified;
determining, using a processor in communication with the computer readable memory, a residual signal in each of the images of the additional grayscale image set based on a combination of images in the additional grayscale image set;
subtracting, using the processor, the residual signal of each image of the set of additional grayscale images from each image of the set of additional grayscale images; and
identifying, using the processor, a defect in the wafer based on differences between the sets of grayscale images;
A method for identifying defects in a wafer by an inspection tool, further comprising: An improved inspection tool system comprising:
control processor;
an electronic image capture device in electronic communication with the control processor;
a plurality of light emitting diodes in electronic communication with the control processor, each light emitting diode configured to emit light of a different wavelength;
a computer readable memory in electronic communication with the image capture device; and
an analysis processor in electronic communication with the computer readable memory
including,
The control processor,
instruct the plurality of light emitting diodes to illuminate at least a portion of the wafer with blue wavelength light and to capture a first grayscale image;
instruct the plurality of light emitting diodes to illuminate at least a portion of the wafer with red wavelength light and capture a second grayscale image;
instruct the plurality of light emitting diodes to illuminate at least a portion of the wafer with green wavelength light and to capture a third grayscale image;
instruct the electronic image capture device to capture a set of grayscale images of a wafer, each image of the set being captured while at least a portion of the wafer is illuminated by the plurality of light emitting diodes;
to store the set of grayscale images in the computer readable memory;
composed,
The analysis processor,
determine a residual signal in each image of the set of grayscale images retrieved from the computer readable memory based on a combination of images in the set of grayscale images;
subtracting the residual signal of each image of the set of grayscale images from each image of the set of grayscale images;
to identify defects in the wafer based on the subtracted set of grayscale images;
An improved inspection tool system comprising: 12. The method of claim 11,
The control processor is further configured to instruct the plurality of light emitting diodes to illuminate at least a portion of the wafer with a combination of blue, red, and green wavelength light and capture an additional grayscale image under the combined light. phosphorus, an improved inspection tool system. 12. The method of claim 11,
and an analog-to-digital converter configured to convert the set of grayscale images for storage in the computer-readable memory. 12. The method of claim 11,
The analysis processor,
using the analysis processor to build a rigorous mathematical model of defect detection using an inspection tool;
determining, using the analysis processor, one or more model parameters based on a known set of standard grayscale images;
constructing, using the analysis processor, a model of the wafer using the one or more model parameters, the model being based on design values or previously measured values;
predict grayscale signals by using the model of the wafer and the rigorous mathematical model;
adjusting one or more parameters of the model of the wafer until a best match is found between the predicted grayscale signals and the measured grayscale signals from the wafer;
reporting, using the analysis processor, the one or more parameters corresponding to the models for which the best match was found as measured sample parameters;
calculating, using the analysis processor, a residual signal based on a difference between the predicted grayscale and the measured grayscale on the wafer;
storing the calculated residual signal in the computer readable memory for future fault detection,
and determining a residual signal in each of the images of the set of grayscale images. 12. The method of claim 11,
wherein the analysis processor is further configured to import wafer information from the computer readable memory and determine a residual signal in respective images of the set of grayscale images based on the imported wafer information. tool system. 16. The method of claim 15,
wherein the wafer information is in GDSII format. 12. The method of claim 11,
the control processor is further configured to instruct the electronic image capture device to capture an additional set of grayscale images of a wafer after the wafer has been modified;
The analysis processor also comprises:
determine a residual signal in each image of the additional grayscale image set based on a combination of the images in the additional grayscale image set;
subtract the residual signal of each image in the additional grayscale image set from each image in the additional grayscale image set;
to identify defects in the wafer based on differences between the sets of grayscale images;
An improved inspection tool system comprising:

Images